Nvidia Senior ML Software Engineer 

United States, California 

282359416

If you care deeply about software craftsmanship, maintainability, and performance—and have hands-on experience building ML systems—this role is for you.

What you’ll be doing:

• Develop and maintain high-quality, modular, and well-tested Python code for large-scale ML infrastructure. Seehttps://github.com/nvidia-cosmos

• Design and optimize post-training, inference, and data processing pipelines used by ground breaking ML models.

• Collaborate with research and product teams to bring ML systems from prototype to production.

• Contribute to open-source projects and build internal tools that enable scalable AI experimentation.

• Improve performance, reliability, and observability of large distributed systems.

• Mentor and support teammates through design reviews, code reviews, and collaborative development.

What we need to see:

• Expert-level proficiency in Python and a track record of delivering production-quality software.

• Strong experience with PyTorch (or similar frameworks such as JAX or TensorFlow), especially in real-world applications.

• Deep understanding of ML system design, training loops, data loaders, evaluation, and model serving.

• Familiarity with containerization, CI/CD, and maintaining in production environments

• Comfortable working with large codebases, building reusable libraries, and writing documentation and tests.

• BSc degree or equivalent experience in Computer Science, Engineering, or a related field is preferred.

• 5+ years of relevant software development experience

Ways to stand out in the crowd:

• Contributions to open-source ML or Python infrastructure projects.

• Background in scaling ML training and inference systems across GPUs as well as experience building libraries that wrap or extend PyTorch functionality.

• Prior exposure to multimodal models or simulation environments (vision, language, audio).

• Familiarity with NVIDIA’s GPU compute stack or high-performance computing clusters.

• Experience with distributed computing tools like DDP, FSDP, ZeRO, or Ray.